I'd like to light up a bunch of LEDs (64, 100, 200,300, or why not, 1000 LEDs...). I suppose i should multiplex them.. but I want to do this without using a Peggy or even a microcontroller. I just want the LEDs to stay on, no fancy animations, blinking or pulsing. Is there an IC to do this ?

If you just want them to stay on, there's no need to multiplex them-- just hard-wire them all in a great big grid-- 100 columns connected to anodes, 100 rows, where each led in each row has cathode to (its own) resistor to ground.

If you really want to multiplex them, you can do it with a 555 counter as an oscillator, driving decade counters (as in the "analog" Larson Scanner), or more optimally, where the ripple counters drive demultiplexers (like the ones in the Peggy 2). Of course, then you'll need transistors to drive the rows, and you could share resistors between the LEDs in a given column (if they're the same color).

Could there be some sort of energy efficiency lighting one small set at a time? Would it outweigh the added circuitry? Or maybe it could be helpful in lowering power supply requirements?

I'm just wondering myself, now.

>Could there be some sort of energy efficiency lighting one small set at a time?

Probably not... but multiplexing would presumably give you more flexibility in how to drive the matrix when it's all built. Without multiplexing, the best that you can do is to go back and change all the resistors.

If you just want them to stay on, there's no need to multiplex them-- just hard-wire them all in a great big grid-- 100 columns connected to anodes, 100 rows, where each led in each row has cathode to (its own) resistor to ground.

I wonder if that setup would make the LEDs be dim and not light up at their full capacity (when lighting up LOTS of LEDs off a 12v power supply) ? Is there an schematic somewhere for this so I can give it a try ? I've always found this confusing... If I have a 12v power supply and let's say, 12 LEDs that drop 1v each, I can't even light up a 13th LED, isn't that right ? And most LEDs drop more than 1v each.. that's why I thought about multiplexing.

Even though I have some 555's lying around I won't even ask anything now because I'm just too clueless atm.. and I don't have this decade conters.

I do have a MAX7219 here, but I think it only works with a microcontroller... it does have a test mode I can't make work - yet. OMG hehe

All I've done so far was light up 15 LEDs to a power supply with no resistors (ouch).

All of this can be a little bit frustrating sometimes...

Any help: tremendously welcome!

>I wonder if that setup would make the LEDs be dim and not light up at their full capacity

No... I was imagining that you would drive each at full power.

When I discussed doing this without multiplexing, I meant that you could wire up each LED independently. For example, let's say that you make a grid of 32 vertical wires, each connected to 5 V, overlaid with 32 horizontal wires connected to ground. You space apart the vertical and horizontal wires so that they don't touch. At each grid intersection point, you add one LED in series with one resistor between +5 and ground.

Using series combinations of LEDs-- as you're asking about --is also possible, but slightly more complex to understand.

If you wanted to do this with 12 V, running seven red LEDs that take 1.7 V each (none get to full brightness at 1 V), you could do that directly, hooking the seven LEDs in series between +12 and ground.

If you're worried about getting "full bright" out of LEDs, multiplexing wouldn't work anyway. For every LED (or row, column, section) being lit, the rest will be off. The more segments to cycle through, the longer each would be off and the dimmer the overall brightness would be.

Running them all hardwired in parallel would only dim them if you're out-drawing what current the power supply can give.

100 columns connected to anodes, 100 rows, where each led in each row has cathode to (its own) resistor to ground.

Let's just say I want to light up 30 LEDs.

Er... I *think* I got it right but I'm not into losing 30 LEDs every time I try something like this... I've made a quick sketch of what I think is right..

http://www.flickr.com/photos/xdgx/4580409886/sizes/o/

If the image is right I can find out the resistor values by myself... the main problem here is the hook-up.

Is that right ?
Thank you

Sure, that would work. But (1) it's not multiplexing-- just driving the LEDs separately and (2) you probably want to use something like 4-5V, not 12 V. Otherwise, you need *big* resistors and they'll get *hot*.

On this schematic the resistors go to the cathodes (common cathode?)
I usually wire the resistor to the anode ( long LED lead ) when driving 1 LED on an arduino pin (or 4 LEDs w/o resistors).
Why is that ?

The resistor doesn't care if it's before or after the LED. You can pick either way for convenience.

The resistor can be connected to the cathode *or* anode ?

Yes, you're just limiting the total amount of juice going through the LED either way. As long as there's nowhere else to go, all the current might want to race through the LED, but since it has to get through the resistor afterwards it causes a traffic jam that slows down everybody.

Great! I think I got it now and made 15 3.2v blue LEDs light up with a 6v battery.

Now I'm going for 100 LEDs.
Can I light them up using the same technique, but hooking them somehow to my arduino so I can make them all blink/fade/respond to a sensor ?

AFAIK 100 3.2v blue LEDs can't be hooked up to an arduino without a trick as there's not enough power coming out of the pins. I think a way to do it is to provide another power source for the LEDs and some kind of 'data' connection for them to blink or fade using a simple sketch.

Hope I can get an answer, thanks!

If you have the LEDs wired up as in your prior diagram (only using 6 V instead of 12), then you can use a PNP transistor to control them. But, doing so is not necessarily trivial since the Arduino outputs do not go as high as 6 V. You'll also have to watch the current rating on the transistor, its saturation voltage, and so forth.